Dynamic interference effect evaluation method based on neural network

Abstract

The invention relates to a dynamic interference effect evaluation method based on a neural network, belongs to the technical field of simulation, and solves the problems that in the prior art, calculation is complex, efficiency is low, and process evaluation is not considered. The method comprises the following steps: collecting interference effect influence factor data of a historical simulation process of a simulation system; calculating the interference effect based on the performance of the detection equipment of the simulation system before and after interference; the interference effect influence factor data and the corresponding interference effect form a training sample set; constructing a neural network model, and training the constructed neural network model based on the training sample set to obtain a corresponding interference effect evaluation model; and inputting to-be-evaluated interference effect influence factor data into the interference effect evaluation model to obtain a corresponding interference effect evaluation result. And rapid and relatively accurate interference effect evaluation is realized.

Description

Technical Field

[0001] The present invention relates to the field of simulation technology, and in particular to a dynamic interference effect evaluation method based on a neural network. Background Art

[0002] In the simulation system, the detection equipment searches and tracks the target, while the jammer is used to interfere with the detection equipment. The jamming effect is an important performance indicator used to measure the results. By quantitatively analyzing the jamming effect and comparing the jamming performance under different strategies, it will be helpful for the analysis and selection of jamming countermeasure strategies. According to the evaluation results, the strategy can be adjusted to optimize the jamming countermeasure performance and put oneself in a favorable position for confrontation. At the same time, it has very important guiding significance for experimental design and the development of jamming countermeasure equipment.

[0003] In the existing simulation system, the interference effect evaluation is obtained by evaluating the interference related indicators using hierarchical indicator calculation methods such as the hierarchical analysis method. The calculation is complex and inefficient, and process evaluation is not considered. Summary of the invention

[0004] In view of the above analysis, an embodiment of the present invention aims to provide a dynamic interference effect evaluation method based on a neural network to solve the problem that the existing method is complex in calculation, inefficient and does not consider process evaluation.

[0005] On the one hand, an embodiment of the present invention provides a method for evaluating dynamic interference effects based on a neural network, comprising the following steps:

[0006] Collecting interference effect influencing factor data of the historical simulation process of the simulation system; calculating the interference effect based on the performance of the detection equipment of the simulation system before and after the interference; the interference effect influencing factor data and the corresponding interference effect constitute a training sample set;

[0007] Constructing a neural network model, and training the constructed neural network model based on the training sample set to obtain a corresponding interference effect evaluation model;

[0008] The interference effect influencing factor data to be evaluated is input into the interference effect evaluation model to obtain the corresponding interference effect evaluation result.

[0009] Based on the further improvement of the above method, the interference effect is calculated based on the performance of the detection equipment of the simulation system before and after the interference, including:

[0010] Calculate a first interference result based on a target search time ratio and a target tracking error ratio of a detection device of a simulation system before and after interference;

[0011] Calculating the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect, and calculating the second interference result based on the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect;

[0012] The interference effect is calculated based on the first interference result and the second interference result.

[0013] Based on the further improvement of the above method, the following formula is used to calculate the training loss of the neural network model:

[0014]

[0015] Among them, ω i represents the weight of the i-th time step, y i represents the true value of the interference effect at the i-th time step, represents the predicted value of the interference effect at the i-th time step, T represents the number of time steps, δ represents the threshold of the prediction error, γ represents the penalty coefficient, and λ represents the decay rate.

[0016] Based on the further improvement of the above method, the weight of the i-th time step is calculated using the following formula:

[0017]

[0018] Among them, k and α are adjustment parameters.

[0019] Based on the further improvement of the above method, the first interference result is calculated based on the target search time ratio and target tracking error ratio of the detection device of the simulation system before and after the interference, including:

[0020] Calculate the ratio of the time it takes the detection device to search for the target after the interference to the time it takes the detection device to search for the target before the interference and perform normalized mapping to obtain the target search time ratio K st ;

[0021] Calculate the ratio of the error of the detection equipment tracking the target after the interference to the error of the detection equipment searching the target before the interference and perform normalized mapping to obtain the target tracking error ratio K te ;

[0022] Based on the formula E1=α1×K st +α2×K te The first interference result is calculated, where α1 and α2 represent weight parameters.

[0023] Based on the further improvement of the above method, the second interference result E is calculated by the following formula based on the time influence effect, space influence effect, frequency influence effect and energy influence effect: j :

[0024]

[0025] R SNR / real =-10lg(e js )-10lg(e jf )+F SNR (P f ,1-e jp )

[0026]

[0027] Among them, e jt Indicates the time effect, the superscript -1 indicates the inverse function, e js Indicates the spatial influence effect, e jf Indicates the frequency effect, e jp Indicates the energy impact effect, P f represents the false alarm probability, Indicates rounding down, F SNR (P f ,1-e jp ) represents the false alarm probability P f , the detection probability is 1-e jp The signal-to-interference ratio corresponding to the detection characteristic curve of the detection equipment, P d (P f ,R SNR|real ) represents the false alarm probability P f , the input signal-to-noise ratio is R SNR|real , I0(·) represents the indicative function, and x represents the simulation time.

[0028] Based on the further improvement of the above method, the following formula is used to calculate the time impact effect:

[0029]

[0030] Among them, t r1 Indicates the time when the detection signal of the detection device reaches the target, t r2 Indicates the time when the detection signal of the detection device ends, t j Indicates the moment when the jammer starts transmitting the jamming signal.

[0031] Based on the further improvement of the above method, the following formula is used to calculate the spatial impact effect:

[0032]

[0033] Among them, θ j represents the antenna axis direction of the jammer, θ j0.5 represents the half-power beamwidth of the jammer’s antenna, θ r Indicates the antenna axis direction of the detection device, θ r0.5Indicates the half-power beamwidth of the antenna of the detection device.

[0034] Based on the further improvement of the above method, the following formula is used to calculate the frequency influence effect:

[0035]

[0036] Among them, f r1 and f r2 Respectively represent the minimum frequency and maximum frequency of the detection equipment receiver, f j1 and f j2 They respectively represent the minimum frequency and maximum frequency of the jammer's interference signal.

[0037] Based on the further improvement of the above method, the following formula is used to calculate the energy impact effect:

[0038] e jp =1-F Pd (P f ,SNR out )

[0039] Among them, P f Indicates the false alarm probability, SNR out represents the output signal-to-noise ratio, F Pd (P f ,SNR out ) represents the signal-to-interference ratio.

[0040] Compared with the prior art, the present invention collects data on interference effect influencing factors of the historical simulation process of the simulation system; calculates the interference effect based on the performance of the detection equipment of the previous and subsequent simulation systems, thereby constructing a training sample set for training the neural network model; an interference effect evaluation model for predicting the interference effect can be obtained by using the neural network model constructed based on the training sample set; the corresponding interference effect evaluation result can be quickly obtained by inputting the interference effect evaluation model into the data to be evaluated, without the need for complex calculations, thereby improving the efficiency of the interference effect evaluation; and can be used for real-time and dynamic effect evaluation of each step of the confrontation process in the simulation system, which helps the confrontation parties to choose a strategy that is more beneficial to themselves according to the current interference effect, thereby improving the credibility and efficiency of the simulation.

[0041] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are only used for the purpose of illustrating specific embodiments and are not to be considered as limiting the present invention. In the entire drawings, the same reference symbols represent the same components;

[0043] Figure 1 Flow chart of a method for evaluating dynamic interference effects based on a neural network according to an embodiment of the present invention

[0044] Figure 2 Schematic diagram of a mapping function according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0046] A specific embodiment of the present invention discloses a method for evaluating dynamic interference effects based on a neural network, such as Figure 1 As shown, the following steps are included:

[0047] S1. Collect data on interference effect influencing factors during the historical simulation process of the simulation system; calculate the interference effect based on the performance of the detection equipment of the simulation system before and after the interference; the interference effect influencing factor data and the corresponding interference effect constitute a training sample set;

[0048] S2, constructing a neural network model, and training the constructed neural network model based on the training sample set to obtain a corresponding interference effect evaluation model;

[0049] S3. The interference effect influencing factor data to be evaluated is input into the interference effect evaluation model to obtain the corresponding interference effect evaluation result.

[0050] Compared with the prior art, the dynamic interference effect evaluation method based on neural network provided in this embodiment collects interference effect influencing factor data of the historical simulation process of the simulation system; calculates the interference effect based on the performance of the detection equipment of the previous and subsequent simulation systems, thereby constructing a training sample set for training the neural network model; an interference effect evaluation model for predicting the interference effect can be obtained by constructing the neural network model based on the training sample set; the corresponding interference effect evaluation result can be quickly obtained by inputting the data to be evaluated into the interference effect evaluation model without the need for complex calculations, thereby improving the efficiency of the interference effect evaluation; and can be used for real-time and dynamic effect evaluation of each step of the confrontation process in the simulation system, which helps the confrontation parties to choose a strategy that is more beneficial to themselves according to the current interference effect, thereby improving the credibility and efficiency of the simulation.

[0051] During implementation, the jammer of the simulation system can be an interference simulation system. The jammer applies interference signals to the jamming direction detection device as an interference direction, and the detection device, as an anti-interference party, eliminates the anti-interference signals through anti-interference measures. The interference effect in the confrontation process between the detection device and the jammer is mainly determined by the inherent characteristics of the detection device, the anti-interference strategy of the detection device, and the key performance of signal reconstruction. Therefore, the present invention constructs interference effect influencing factors from three dimensions: inherent indicators of the simulation system (referred to as Class I indicators), indicators related to the use strategy (referred to as Class II indicators), and indicators related to signal reconstruction (referred to as Class III indicators).

[0052] Specifically, the inherent indicators include maximum output power (unit: W), instantaneous bandwidth (unit: Hz), target distance simulation accuracy (unit: m), and target angle simulation accuracy (unit: °, interval range).

[0053] The use of strategy mainly refers to the interference and anti-interference methods adopted by the opposing parties during the confrontation process.

[0054] The time range in which the jammer can effectively interfere with the target detection device is called the effective jamming time. Before generating the jamming signal, the jammer needs to detect the various parameters of the target detection device, and then generate the predetermined jamming signal. After the jamming is implemented, it is necessary to continuously detect the changes in the detection device signal, so as to prepare for the next step of jamming. Therefore, the jammer must have a fast enough guidance time to effectively implement the jamming of the target detection device. The factors related to the jamming timing include two aspects: the start time and duration of the jamming.

[0055] Interference airspace factors can be considered as a prerequisite for the success of interference. Only when the interference signal direction is correct can the subsequent other interference factors be guaranteed to play a role.

[0056] The threat space range of the detection device or the interference range required by the jammer is Ω, and the interference range of the jammer is Ω j The overlapping area between the jammer's jamming range and the jamming range required for combat is ΔΩ j , assuming that these ranges are known quantities based on reconnaissance information. Then define an effective interference coverage benefit factor E s , to characterize the factors affecting the interference airspace.

[0057]

[0058] In the above formula, ΔΩ j The value range of is [0,Ω], so E s The value ranges from 0 to 1. The start time and duration of interference application and the interference coverage efficiency factor are related indicators of signal reconstruction.

[0059] The power factor plays a vital role in the detection equipment confrontation. As the detection party, the output power of the transmitter will directly affect the working performance and anti-interference ability of the detection equipment. For the jammer, the power influencing factor is characterized by the effective radiated power, which refers to the transmit power P of the jammer. J and the transmitting antenna gain G J The power factor is an inherent indicator of the simulation system.

[0060] The interference types of detection equipment generally include suppressive interference and deceptive interference. The interference type of the detection equipment must match the working state, working system, detection equipment function and anti-interference measures of the detection equipment to achieve a better interference effect. The choice of interference type is also closely related to the working system of the detection equipment and the corresponding anti-interference measures. During implementation, the interference benefit factor E5 is used to indicate the impact of the interference type on the interference effect of the detection equipment. In general, the interference type benefit factor cannot be quantitatively described by a fixed value or mathematical expression. Based on the matching relationship between the interference type and the working state, working system and anti-interference measures of the detection equipment described above, combined with the expert prior knowledge, a value between 0 and 1 can be subjectively given as the size of the interference style benefit factor.

[0061] In the process of signal reconstruction, in addition to inherent indicators and indicators related to the use of strategies, the decisive factors of the interference effect can be summarized into the following main aspects:

[0062] (1) The frequency and direction of the interference are aligned with the interference detection equipment;

[0063] (2) The relative size of the interference power;

[0064] (3) Countermeasures against electromagnetic, meteorological and geographical factors in space.

[0065] After collecting the interference-related factors of each step of the historical confrontation process, it is necessary to calculate the corresponding interference effect value.

[0066] Calculate the interference effect based on the performance of the detection equipment of the simulated system before and after the interference, including:

[0067] S11, calculating a first interference result based on a target search time ratio and a target tracking error ratio of a detection device of a simulation system before and after interference;

[0068] S12, calculating the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect, and calculating the second interference result based on the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect;

[0069] S13. Calculate an interference effect based on the first interference result and the second interference result.

[0070] In order to measure the interference effect more comprehensively and accurately, the present invention evaluates the interference effect from two aspects: the interference result and the interference process, namely, the first interference result and the second interference result.

[0071] The problem to be solved in interference effect measurement is how to measure the specific interference effect based on the changes in the working state of the detection equipment before and after the interference in the experiment. In order to achieve quantitative measurement of interference effect, it is necessary to reasonably select measurement indicators, and the selection of indicators should meet the following principles:

[0072] 1) Although the interference effect is related to many factors, since it reflects the degree of influence of the interference measures on the detection equipment, it is necessary to select indicators that can directly reflect the changes in the working performance of the detection equipment before and after the interference;

[0073] 2) The interference effect is the result of the combined influence of multiple factors, so the selected indicators should be able to comprehensively reflect as many influencing factors as possible and be able to change with the changes of influencing factors;

[0074] 3) The indicator should be easy to measure and can be conveniently measured on the detection equipment;

[0075] 4) In view of the active interference faced by the detection equipment, the measurement indicators are required to reflect both the interference degree of the search segment and the interference degree of the tracking segment, and both the effect of masking interference and the effect of deceptive interference.

[0076] Based on the above analysis, combined with the "time criterion" and "efficiency criterion", it is proposed to use the search time ratio and tracking error ratio as indicators to measure the active interference effect of the detection equipment. Specifically, the first interference result is calculated based on the target search time ratio and target tracking error ratio of the detection equipment of the simulation system before and after the interference, including:

[0077] Calculate the ratio of the time it takes the detection device to search for the target after the interference to the time it takes the detection device to search for the target before the interference and perform normalized mapping to obtain the target search time ratio K st ;

[0078] Calculate the ratio of the error of the detection equipment tracking the target after the interference to the error of the detection equipment searching the target before the interference and perform normalized mapping to obtain the target tracking error ratio K te ;

[0079] Based on the formula E1=α1×K st +α2×K te The first interference result is calculated, where α1 and α2 represent weight parameters.

[0080] During implementation, in order to obtain the interference effect of the interference result angle, it is necessary to comprehensively process two indicators. First, calculate the ratio of the time for the detection device to search for the target after interference to the time for the detection device to search for the target before interference, and then normalize it, that is, select an appropriate function according to the characteristics of the indicator to map it to the interval [0,1], so as to obtain the target search time ratio K st 。

[0081] The search time ratio K st The larger it is, the better the interference effect. Therefore, the Gaussian lower-type normal distribution function in fuzzy mathematics is used to map them to the interval [0,1]. The specific function is as follows.

[0082]

[0083] Among them, x represents the ratio of the time for the detection device to search for the target after interference to the time for the detection device to search for the target before interference before normalization, and f(x) represents the normalization function. Figure 2 The mapping function graph is given. When x < L, f(x) = 0. Therefore, L is the upper limit value of the measurement index corresponding to ineffective interference (that is, the ratio of search time). When x > L, the interference starts to take effect. When x > H, f(x) approaches 1. H corresponds to the lower limit value of the measurement index when the detection device completely fails. ε = (H - L) / 3. The interval [L, H] corresponds to the range from ineffective interference to the most severe interference.

[0084] During implementation, the time for the detection device to search for the target before interference is the normal search time of the detection device, and the time for the detection device to search for the target after interference is the time exceeding the normal search time after interference.

[0085] For the search time ratio K st , L is determined by the longest search time that can be allowed without affecting the normal operation of the detection device, and H is determined by the shortest search time in the case of the detection device failure.

[0086] During implementation, the error of the detection device tracking the target before interference is the error of the detection device tracking the target under normal conditions, and the error of the detection device tracking the target after interference is the difference between the tracking error after interference and the tracking error under normal conditions.

[0087] After calculating the ratio of the error of the detection device tracking the target after interference to the error of the detection device searching for the target before interference and performing normalization mapping on it, the target tracking error ratio K te is obtained. During implementation, the normalization function can be in the same form as formula (2), but L and H in it need to be determined according to the tracking error of the detection device. For the tracking error ratio K te , usually L = 2 is set in the field. H is determined by the maximum tracking error that the detection device can allow.

[0088] Then, the mapped indicators are weighted and summed, that is, according to E1 = α1 × K st +α2×K te The first interference result is calculated, where α1 and α2 represent weight parameters. For the detection device, being in the search state means that the detection device cannot track the real target at all, so α1>α2.

[0089] The interference effect of the interference process can be evaluated from the perspectives of time, space, frequency and energy, and the time influence effect, space influence effect, frequency influence effect and energy influence effect can be calculated respectively. The second interference result is calculated based on the time influence effect, space influence effect, frequency influence effect and energy influence effect.

[0090] Specifically, the time impact effect is calculated using the following formula:

[0091]

[0092] Among them, t r1 Indicates the time when the detection signal of the detection device reaches the target, t r2 Indicates the time when the detection signal of the detection device ends, t j Indicates the moment when the jammer starts to transmit the jamming signal. jt It reflects the degree of suppression of the detection signal of the detection equipment by the jammer in time. The time influence effect jt The larger it is, the more time the interference signal covers the detection signal in the time domain, and the better the interference effect is.

[0093] Specifically, the following formula is used to calculate the spatial impact effect:

[0094]

[0095] Among them, θ j represents the antenna axis direction of the jammer, θ j0.5 represents the half-power beamwidth of the jammer’s antenna, θ r Indicates the antenna axis direction of the detection device, θ r0.5 Indicates the half-power beamwidth of the antenna of the detection device. js It reflects the degree of spatial alignment between the jammer antenna and the detection equipment receiving antenna. js The larger it is, the higher the degree of alignment, the greater the antenna gain, and the better the interference effect.

[0096] Specifically, the frequency impact effect is calculated using the following formula:

[0097]

[0098] Among them, fr1 and f r2 Respectively represent the minimum frequency and maximum frequency of the detection equipment receiver, f j1 and f j2 They respectively represent the minimum frequency and maximum frequency of the jammer's interference signal.

[0099] Frequency Effect jf It reflects the degree to which the interference signal frequency enters the detection equipment receiver, e jf The larger it is, the more interference signals enter the detection device receiver and the better the interference effect will be.

[0100] The energy impact effect is calculated using the following formula:

[0101] e jp =1-F Pd (P f ,SNR out ) (6)

[0102] Among them, P f Indicates the false alarm probability, SNR out represents the output signal-to-noise ratio, F Pd (P f ,SNR out ) represents the signal-to-interference ratio.

[0103] The energy impact effect reflects the probability of false alarm of the detection equipment. The greater the energy impact effect, the greater the probability of false alarm and the better the interference effect.

[0104] The second interference result E is calculated based on the time influence effect, space influence effect, frequency influence effect and energy influence effect using the following formula j :

[0105]

[0106] R SNR / real =-10lg(e js )-10lg(e jf )+F SNR (P f ,1-e jp ) (8)

[0107]

[0108] Among them, e jt Indicates the time effect, the superscript -1 indicates the inverse function, e js Indicates the spatial influence effect, e jf Indicates the frequency effect, e jp Indicates the energy impact effect, P f represents the false alarm probability, Indicates rounding down, F SNR (P f ,1-e jp ) represents the false alarm probability P f , the detection probability is 1-e jp The signal-to-interference ratio corresponding to the detection characteristic curve of the detection equipment, P d (P f ,R SNR|real ) represents the false alarm probability P f , the input signal-to-noise ratio is R SNR|real , I0(·) represents the indicative function, and x represents the simulation time.

[0109] Among them, R SNR|real is the power ratio of the actual target signal to the interference signal, F SNR (P f ,1-e jp ) represents the false alarm probability P f , the detection probability is 1-e jp The signal-to-interference ratio corresponding to the detection characteristic curve of the detection equipment at that time. Its value can be obtained by inverting the detection characteristic function of the detection equipment.

[0110] After obtaining the first interference result and the second interference result, the total interference effect can be calculated by weighted summing the first interference result and the second interference result. After calculating the interference factor data and the corresponding interference effect of each step of confrontation, a training sample set is constructed. Since the interference effect of the confrontation process is time-dependent, when implementing, the data is sorted according to the time sequence of the confrontation process, and the data of consecutive T steps is extracted as a sample to construct a training sample set. When implementing, a part of the samples can be drawn from the training sample set as a verification set to verify the effect of the model.

[0111] During implementation, the constructed neural network model is a recurrent neural network, including an input layer, a multi-layer recurrent layer, and an output layer;

[0112] The input layer is used to input sample data, and the input data is transmitted to the multi-layer circulation layer;

[0113] Each recurrent layer includes a plurality of gated recurrent units connected in sequence, and is used to extract hidden features of input data;

[0114] The output layer includes a plurality of output units, and the output units correspond one-to-one to the gated recurrent units of the last recurrent layer, and are used to predict the interference effect value according to the output of the gated recurrent unit.

[0115] By adopting a recurrent neural network, the temporal dependency characteristics of the adversarial process are mined, so that the trained model can more accurately predict the interference effect.

[0116] When implemented, the training loss of the neural network model is calculated using the following formula:

[0117]

[0118] Among them, ω i represents the weight of the i-th time step, y i represents the true value of the interference effect at the i-th time step, represents the predicted value of the interference effect at the i-th time step, T represents the number of time steps, δ represents the threshold of the prediction error, γ represents the penalty coefficient, and λ represents the decay rate.

[0119] In order to improve the robustness of the training loss to outliers, when the loss error is greater than the threshold prediction error, the penalty for larger errors is increased exponentially. As the error increases, the penalty term increases rapidly, thereby imposing a greater penalty on large errors.

[0120] During implementation, in order to improve the efficiency of training, different weights are assigned to each time step.

[0121] Specifically, the weight of the i-th time step is calculated using the following formula:

[0122]

[0123]

[0124] Among them, k and α are adjustment parameters.

[0125] When the training loss reaches a preset accuracy or the training reaches a preset number of iterations, the interference effect evaluation model is obtained.

[0126] For the interference effect influencing factor data to be evaluated, it is input into the interference effect evaluation model to obtain the corresponding interference effect evaluation result. For example, if the interference effect of the current confrontation is evaluated, the interference effect influencing factor data of the current confrontation and the previous T-1 confrontation can be extracted and input into the interference effect evaluation model to obtain the corresponding interference effect evaluation result.

[0127] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0128] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A dynamic interference effect evaluation method based on neural network, characterized in that: The following steps are involved: Collect data on interference effect influencing factors of the historical simulation process of the simulation system; The interference effect is calculated based on the performance of the detection equipment of the simulation system before and after the interference; the interference effect influencing factor data and the corresponding interference effect constitute a training sample set; Constructing a neural network model, and training the constructed neural network model based on the training sample set to obtain a corresponding interference effect evaluation model; The interference effect influencing factor data to be evaluated is input into the interference effect evaluation model to obtain the corresponding interference effect evaluation result.

2. The method for dynamic interference effect evaluation based on neural network according to claim 1, characterized in that: Calculate the interference effect based on the performance of the detection equipment of the simulated system before and after the interference, including: Calculate a first interference result based on a target search time ratio and a target tracking error ratio of a detection device of a simulation system before and after interference; Calculating the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect, and calculating the second interference result based on the time influence effect, the space influence effect, the frequency influence effect and the energy influence effect; The interference effect is calculated based on the first interference result and the second interference result.

3. The method for dynamic interference effect evaluation based on neural network according to claim 1, characterized in that: The training loss of the neural network model is calculated using the following formula: Among them, ω i represents the weight of the i-th time step, y i represents the true value of the interference effect at the i-th time step, represents the predicted value of the interference effect at the i-th time step, T represents the number of time steps, δ represents the threshold of the prediction error, γ represents the penalty coefficient, and λ represents the decay rate.

4. The method for dynamic interference effect evaluation based on neural network according to claim 3 is characterized in that: The weight of the i-th time step is calculated using the following formula: Among them, k and α are adjustment parameters.

5. The method for dynamic interference effect evaluation based on neural network according to claim 2, characterized in that: Calculating the first interference result based on the target search time ratio and target tracking error ratio of the detection device of the simulation system before and after the interference includes: Calculate the ratio of the time it takes the detection device to search for the target after the interference to the time it takes the detection device to search for the target before the interference and perform normalized mapping to obtain the target search time ratio K st ; Calculate the ratio of the error of the detection equipment tracking the target after the interference to the error of the detection equipment searching the target before the interference and perform normalized mapping to obtain the target tracking error ratio K te ; Based on the formula E1=α1×K st +α2×K te The first interference result is calculated, where α1 and α2 represent weight parameters.

6. The dynamic interference effect evaluation method based on neural network according to claim 2 is characterized in that: The second interference result E is calculated based on the time influence effect, space influence effect, frequency influence effect and energy influence effect using the following formula j : R SNR / real =-10lg(e js )-10lg(e jf )+F SNR (P f ,1-e jp ) Among them, e jt Indicates the time effect, the superscript -1 indicates the inverse function, e js Indicates the spatial influence effect, e jf Indicates the frequency effect, e jp Indicates the energy impact effect, P f represents the false alarm probability, Indicates rounding down, F SNR (P f ,1-e jp ) represents the false alarm probability P f , the detection probability is 1-e jp The signal-to-interference ratio corresponding to the detection characteristic curve of the detection equipment, P d (P f ,R SNR|real ) represents the false alarm probability P f , the input signal-to-noise ratio is R SNR|real , I0(·) represents the indicative function, and x represents the simulation time.

7. The dynamic interference effect evaluation method based on neural network according to claim 2 is characterized in that: The time impact is calculated using the following formula: Among them, t r1 Indicates the time when the detection signal of the detection device reaches the target, t r2 Indicates the time when the detection signal of the detection device ends, t j Indicates the moment when the jammer starts transmitting the jamming signal.

8. The dynamic interference effect evaluation method based on neural network according to claim 2 is characterized in that: The spatial impact effect is calculated using the following formula: Among them, θ j represents the antenna axis direction of the jammer, θ j0.5 represents the half-power beamwidth of the jammer’s antenna, θ r Indicates the antenna axis direction of the detection device, θ r0.5 Indicates the half-power beamwidth of the antenna of the detection device.

9. The dynamic interference effect evaluation method based on neural network according to claim 2 is characterized in that: The frequency effect is calculated using the following formula: Among them, f r1 and f r2 Respectively represent the minimum frequency and maximum frequency of the detection equipment receiver, f j1 and f j2 They respectively represent the minimum frequency and maximum frequency of the jammer's interference signal.

10. The method for dynamic interference effect evaluation based on neural network according to claim 2, characterized in that: The energy impact effect is calculated using the following formula: e jp =1-F Pd (P f ,SNR out ) Among them, P f Indicates the false alarm probability, SNR out represents the output signal-to-noise ratio, F Pd (P f ,SNR out ) represents the signal-to-interference ratio.

Images